The headline is broadly defensible, but the qualifications matter. Effect sizes vary by population, the strongest claims rest on shorter trials, and credible voices push back on how it's typically framed.
we were going to need to use AI or machine learning um because um we had to learn these complex associations and patterns that um you know no human being could curate thousands and of um you know different mutational profiles and try to find the common signals and so on
Every Sunday: the week’s new conflicts and verdict changes — and nothing else.
Native comments, Twitter mentions, and Reddit threads about this claim — surfaced together so the conversation isn't fragmented across platforms.
Bookmarking — the dossier-vs-overview split is the right call. Most of the time I want overview; sometimes I want receipts.
Would love a "what would change this verdict" RSS feed. Sign me up if it exists.
we were going to need to use AI or machine learning um because um we had to learn these complex associations and patterns that um you know no human being could curate thousands and of um you know different mutational profiles and try to find the common signals and so on